Académique Documents
Professionnel Documents
Culture Documents
Background: Much of our knowledge of the role of ge- phobias. All 5 phobia subtypes aggregate within twin-
netic factors in the etiology of phobias comes from one pairs. This aggregation is due largely or solely to genetic
population-based sample of female twins. We examined factors with heritability of liabilities ranging from 25%
the sources of individual differences in the risks for pho- to 37%. Multivariate analysis revealed a common ge-
bias and their associated irrational fears in male twins. netic factor, genetic factors specific to each subtype, and
a common familial-environmental factor.
Methods: In personal interviews with both members of
1198 male-male twin pairs (707 monozygotic [MZ] and Conclusions: In male subjects, genetic risk factors,
491 dizygotic [DZ]) ascertained from a population- which are partially common across all subtypes and
based registry, we assessed the lifetime history of agora- partially subtype specific, play a moderate role in the
phobia and social, animal, situational, and blood/injury etiology of phobias and their associated irrational fears.
phobias as well as their associated irrational fears. Twin Family environment probably has an impact on risk for
resemblance was assessed by means of probandwise con- agoraphobia and social phobia. The genetic liability to
cordance, odds ratios, tetrachoric correlations, and uni- blood/injury phobias is not distinct from those of the
variate and multivariate biometrical model fitting. more typical phobias.
F
AMILY STUDIES suggest that measurement error suggested that the re-
phobias are familial,1-6 but sults for these phobia subtypes were simi-
cannot clarify the origin of lar to those for the others (ie, the familial
this family resemblance. Twin transmission due solely to genetic factors).
studies of self-reported fears Third, high levels of comorbidity were seen
consistently have suggested a significant role between the phobia subtypes.15 Contrary to
for genetic factors.7-11 By contrast, twin stud- results from family studies,4 the best-
ies of clinically defined phobias have, with fitting multivariate model suggested com-
one exception, suffered from very small mon genetic and individual-specific envi-
samples sizes with resultant low power to ronmental factors that influenced risk for
discriminate alternative models of familial all phobia subtypes, as well as genetic and
transmission.12-14 This exception is our pre- environmental factors unique to each sub-
vious study of phobias in female-female type. However, these analyses did not in-
twins from the Virginia Twin Registry.15-17 clude BiP, which was assessed at a later wave.
These reports examined 5 types of pho- Blood/injury phobia differs from typical pho-
bias—agoraphobia (AgP), social phobia bias,18 where exposure to phobic stimuli (eg,
From the Virginia Institute for (SoP), animal phobia (AnP), situational pho- snakes, heights, and public speaking) usu-
Psychiatric and Behavioral bia (SiP), and blood/injury phobia (BiP)— ally produces increased sympathetic activ-
Genetics (Drs Kendler, and reached the following conclusions. First, ity (eg, tachycardia, increased blood pres-
Prescott, and Neale and all phobia subtypes are moderately famil- sure, sweating, and flushing). By contrast,
Mr Myers) and the ial. Second, familial aggregation of AgP, SoP, in individuals with BiP, exposure to pho-
Departments of Psychiatry and AnP is due to genetic factors. For SiP bic stimuli (eg, needles and blood) usually
(Drs Kendler, Prescott, and
Neale and Mr Myers) and
and BiP, results based on our first assess- increases parasympathetic activity (eg, bra-
Human Genetics (Drs Kendler ment15,16 suggested that twin resemblance dycardia, hypotension, pallor, and fainting).
and Neale), Medical College might result from familial-environmental Given these differences, would the genetic
of Virginia of Virginia factors. However, when our analyses in- and environmental risk factors for BiP be dis-
Commonwealth University, cluded a second assessment interview,17 the tinct from those for the other phobia sub-
Richmond. added power obtained through control of types?
In this report, using our recently studied sample of in the direction opposite that predicted. Controlling for
male-male twin pairs ascertained from the same regis- zygosity, twin pairs with more similar childhood environ-
try, we address the following questions: ments were significantly less likely to be concordant for
AnP (x21 =5.55; P=.02) and SiP (x21 =4.10; P=.04). Those
1. What are the sources of individual differences in
in frequent adult contact were significantly more likely to
risk for the 5 phobia subtypes in male twins?
be concordant for SiP (x21 =5.27; P=.02).
2. Do irrational fears without impairment reflect a
Short-term test-retest reliability for irrational fears and
milder form of the same liability dimensions as classic
phobias, as assessed by the coefficient and the tetrachoric
phobias? If so, will including such fears in the modeling
correlation, were modest to moderate (Table 2). With the
improve power and the ability to discriminate compet-
exception of social fears/SoP, irrational fears were more re-
ing models?
liably reported than phobias. The reliability of the assign-
3. What is the level of comorbidity among individual
ment of a phobia given an irrational fear was particularly
phobia subtypes in this population sample? To what degree
modest. Consistent with our previous results,17 individu-
do genetic and environmental common factors contribute
als may be more reliable at recalling irrational fears than
to the observed patterns of comorbidity, and will these com-
they are at reporting their behavioral consequences.
mon factors have less impact on the liability to BiP than on
the liability to the more typical phobia subtypes?
DESCRIPTION BY INDIVIDUAL FEAR
RESULTS Many more individuals endorsed an irrational fear than
met criteria for an associated phobia (Table 1). How-
TEST FOR BIASES AND RELIABILITY ever, this ratio differed widely. For example, although
46% of individuals who feared being in crowds met cri-
We performed 12 analyses predicting twin concordance teria for AgP, the parallel figures for fears of needles, giv-
for each phobia subtype and any phobias from the simi- ing a speech, and snakes were 18%, 15%, and 14%, re-
larity of their childhood environment and the frequency spectively. Table 1 also presents, within each of the phobia
of adult contact. Three were significant at the 5% level; 2, subtypes, the proportion of individuals who met crite-
ria for phobia based on each specific fear. The most com- sulted from both genetic and familial-environmental fac-
mon impairing fears for each subtype were being in crowds tors. When we tested the simpler AE and CE models in
for AgP, giving a speech for SoP, snakes for AnP, other these 2 phobias, they produced nearly identical fits. Table
high places for SiP, and dentists or hospitals for BiP. 4 presents results for the AE model. The CE model pro-
The most common phobia subtype was SiP, fol- duced the following estimates for c2 and e2: −0.29 and
lowed by SoP, BiP, AnP, and AgP. No significant differ- 0.71, respectively, for AgP, and 0.17 and 0.83, respec-
ences were seen in the prevalence of the phobia subtypes tively, for SoP. The E-only model, by contrast, fit more
or any phobia in MZ vs DZ twins (results not shown). poorly for both phobias. Although AgP and SoP aggre-
gated within twin pairs, we had no power to distinguish
TWIN RESEMBLANCE the degree to which this was due to genetic and/or en-
vironmental mechanisms.
Twin resemblance was seen for all 5 of the phobia sub- For AnP, SiP, BiP, and any phobia, c2 was esti-
types in MZ twins, with the odds ratios (ORs) ranging mated at 0 in the full model, indicating that twin resem-
from 2.30 for SoP to 4.64 for AnP (Table 3). In DZ twins, blance was ascribed entirely to genetic factors. In each
twin resemblance was seen for AgP, SoP, and AnP, but case, the AE model fit somewhat better than the CE
not for SiP or BiP. For all phobia subtypes as well as for model, with estimates of heritability ranging from 0.22
any phobia, the OR and tetrachoric correlations in MZ (for any phobia) to 0.35 (for AnP). The very wide confi-
twins exceeded those seen in DZ twins, suggesting the dence intervals (CIs) on these heritability estimates are
importance of genetic risk factors. noteworthy. Furthermore, using the more rigorous x2
difference test, we could not reject the CE against the
UNIVARIATE MODEL FITTING ACE model at any level approaching statistical signifi-
TO PHOBIAS ONLY cance. Indeed, even the test for familial aggregation of
phobias (the ACE vs E-only model) was significant only
Table 4 shows the results of model fitting for each of for AgP (x22 =6.05; P=.05), AnP (x22 =6.65; P=.04), SiP
the phobia subtypes and for any phobia. For AgP and SoP, (x 2 2 = 7.02; P = .03), and any phobia (x 2 2 = 10.83;
the full ACE model suggested that twin resemblance re- P=.004).
Phobia
Fear† Phobia† Fear-Phobia Dimension‡ (for Those Having Fear)†§
Table 3. Prevalence and Measures of Similarity for Phobias in MZ and DZ Male-Male Twin Pairs*
*MZ indicates monozygotic; DZ, dizygotic; CI, confidence interval; and ASE, asymptotic SE.
†Ellipses indicate results were not presented because multiple threshold model fit poorly.
alone), we found evidence of the impact of shared envi- Individual phobia subtypes and any phobia aggre-
ronment, but only in the form of a common factor. This gated within twin pairs. For AnP, SiP, BiP, and any pho-
common factor significantly affected only AgFP and bia, the best-fit model suggested that familial aggrega-
SoFP. Fourth, the best-fit model contained one com- tion was due solely to genetic factors with modest
mon E factor. That is, some environmental experiences heritabilities. For AgP and SoP, model fitting could not
that were unique to individual twins influence the gen- distinguish between genetic and familial-environmen-
eral risk for fears/phobias. This factor had highest load- tal sources of twin resemblance. For all the phobia sub-
ings on AgFP and SoFP. Finally, substantial specific E types, power was limited, and CIs for the parameter es-
loadings were seen for all subtypes, which would repre- timates were broad.
sent an admixture of measurement error and environ- Many twins in our sample reported irrational fears
mental experiences that predisposed uniquely to indi- without objective impact on their lives. Given our need
vidual phobia subtypes. for greater statistical power to resolve sources of indi-
vidual differences and the low reliability with which in-
COMMENT dividuals recalled whether the fears were impairing, we
applied the multiple-threshold model28 to a trichoto-
We addressed 3 questions about the genetic epidemiol- mous classification of all individuals into unaffected, fear
ogy of irrational fears and phobias in male-male twin pairs only, and phobia. Results from these analyses indicated
from a population-based registry. We examine these ques- that for all the phobia subtypes, the pattern of results
tions in turn. within MZ and DZ twins pairs were consistent with the
Parameter Estimates
*A indicates additive genetic factors; C, family or common environment factors; E, individual specific environment factors; CI, confidence interval; and AIC,
Akaike’s Information Criterion.25 Analyses were based on 1198 complete twin pairs and 544 single twins.
†Degrees of freedom were 2937 for ACE, 2938 for AE and CE, and 2939 for E.
‡For these phobias, we present parameters from the AE model as the best fit, but the CE model also fit nearly as well, results of which are presented in the
“Univariate Model Fitting to Phobias Only” subsection of the “Results” section.
§Best-fit models by means of AIC.
Table 5. Model-Fitting Results for Fears and Phobias in Male-Male Twin Pairs*
Parameter Estimates
*Abbreviations and degrees of freedom for fit of models are given in the first and second footnotes to Table 4. Analyses were based on 1198 complete twin
pairs and 544 single twins. Degrees of freedom for fit of models were as follows: ACE, 2936; AE, 2937; CE, 2937; and E, 2938.
†Best fit model(s) by means of AIC.25
hypothesis that fears were a milder manifestation of the best-fit model for all the phobia subtypes suggested that
same liability dimension that produced clinical pho- familial resemblance was due solely to genetic factors, again
bias. We found similar results in our female twins.17 with modest heritability estimates, but now with consid-
We repeated our analysis using this trichotomous erably smaller CIs.
classification, with results that were substantially clearer We recently fitted multiple threshold models to data
than those obtained for phobias alone. With the in- on fears and phobias from the previously studied female-
creased statistical power, evidence of familial-environ- female pairs17 to which our current results can be com-
mental effects on AgP and SoP disappeared. The pared usefully. We used a measurement model based on
Phobia Agoraphobia Social Phobia Animal Phobia Situational Phobia Blood/Injury Phobia
Agoraphobia 12.78 (8.51-19.18) 5.37 (3.32-8.70) 3.70 (2.40-5.68) 3.44 (2.05-5.78)
Social 0.61 (0.05) 4.59 (3.01-7.01) 3.83 (2.68-5.46) 5.89 (3.99-8.70)
Animal 0.40 (0.06) 0.38 (0.06) 3.81 (2.59-5.59) 5.20 (3.39-7.96)
Situational 0.33 (0.06) 0.36 (0.05) 0.35 (0.05) 4.40 (3.07-6.31)
Blood/injury 0.29 (0.07) 0.45 (0.05) 0.41 (0.06) 0.39 (0.05)
*Data in lower-left triangle are given as tetrachoric correlations (asymptotic SE); in upper-right triangle, odds ratios (95% confidence intervals).
for SoFP, 0.31; for AnFP, 0.27; for SiFP, 0.26; and for 0.47 0.48 – 0.12 0.50
BiFP, 0.32. These results, well within the 95% CIs ob- 0.36 0.22 0.03 0.65
tained in the male sample, suggest that the role of ge- 0.46 – 0.03 0.38
0.33 0.22 0.34 0.46
netic factors in liability to irrational fears and phobias are
probably similar in men and women.
Similar to other community samples,15,30 substan-
tial comorbidity was seen between phobia subtypes. Mul-
tivariate twin modeling is a powerful method to exam- Agoraphobia Social Animal Situational Blood/Injury
ine the contributions of genetic and environmental factors Phobia Phobia Phobia Phobia
and other differences seen in both sexes are substantive netic proportion of reliable variance) would be consid-
in nature or due to stochastic fluctuations in patterns of erably higher than those reported herein.
phobia resemblances in twin pairs.
Our multivariate analyses also permitted us to evalu- Accepted for publication October 10, 2000.
ate the hypothesis that the genetic risk factors for BiP are This work was supported by grants MH/AA-49492 and
distinct from those for the more typical phobia sub- MH-54150 and Research Scientist awards MH-01277 (Dr
types. Our results were inconsistent with this hypoth- Kendler) and MH-01458 (Dr Neale) from the National In-
esis. Indeed, BiP had the highest loading on the com- stitutes of Health, Bethesda, Md. We acknowledge the con-
mon genetic factor, and we could not set that path to 0 tribution of the Virginia Twin Registry, now part of the Mid-
without a substantial deterioration in fit. The more typi- Atlantic Twin Registry, to ascertainment of subjects for this
cal phobias and BiP probably share common early fear study. The Mid-Atlantic Twin Registry, directed by Linda
pathways that then diverge in their outflow to the hy- Corey, PhD, and Lenn Murrelle, PhD, has received support
pothalamus and autonomic pathways.31 These findings from the National Institutes of Health, the Carman Trust
suggest that genetic risk factors for phobias in men act (Richmond, Va), and the WM Keck (Los Angeles, Calif),
largely on the individual differences in the sensitivity of John Templeton (Radnor, Pa), and Robert Wood Johnson
those early fear pathways shared by BiP and the more typi- (Princeton, NJ) Foundations.
cal phobias. These data were collected under the direction of Patsy
These analyses should be considered in the context Waring, Sarah Woltz, MA, and Frank Butera, MS.
of 2 potential methodological limitations. First, our cri- Corresponding author and reprints: Kenneth S. Ken-
teria for phobias differ from those proposed in recent DSM dler, MD, Department of Psychiatry, Medical College of Vir-
editions.32,33 For example, our interviewer-based assess- ginia of Virginia Commonwealth University, 800 E Leigh
ment of impairment, although more objective, may pro- St, PO Box 980126, Richmond, VA 23298-0126.
duce a lower threshold for interference than the self-
report measure used in the DIS. Our definition may also
REFERENCES
be broader than that proposed in DSM-III-R criteria32 in
that we require objective impact on respondent behav-
ior rather than the significant interference with normal 1. Noyes R, Crowe RR, Harris EL, Hamra BJ, McChesney CM, Chaudhry DR. Re-
lationship between panic disorder and agoraphobia. Arch Gen Psychiatry. 1986;
routine in DSM-III-R. However, DSM-III-R would in- 43:227-232.
clude an unreasonable fear that produced marked dis- 2. Fyer AJ, Mannuzza S, Gallops MS, Martin LY, Aaronson C, Gorman JM, Liebo-
tress, whereas we did not. The Epidemiological Catch- witz MR, Klein DF. Familial transmission of simple phobias and fears: a prelimi-
ment Area screened for a more limited number of phobic nary report. Arch Gen Psychiatry. 1990;47:252-256.
stimuli (14 specific fears) and produced a lifetime preva- 3. Fyer AJ, Mannuzza S, Chapman TF, Liebowitz MR, Klein DF. A direct interview
family study of social phobia. Arch Gen Psychiatry. 1993;50:286-293.
lence for any phobia in men of 10.4%,30 about half of what 4. Fyer AJ, Mannuzza S, Chapman TF, Martin LY, Klein DF. Specificity in familial
we found in this study. By contrast, compared with our aggregation of phobic disorders. Arch Gen Psychiatry. 1995;52:564-573.
results, the National Comorbidity Study found much 5. Mannuzza S, Schneier FR, Chapman TF, Liebowitz MR, Klein DF, Fyer AJ. Gen-
higher rates of SoP in men (11.1%)34 and slightly lower eralized social phobia: reliability and validity. Arch Gen Psychiatry. 1995;52:230-
237.
rates of AgP (3.5%).
6. Stein MB, Chartier MJ, Hazen AL, Kozak MV, Tancer ME, Lander S, Furer P, Chu-
Second, a lifetime history of irrational fears and pho- baty D, Walker JR. A direct-interview family study of generalized social phobia.
bias was assessed in this study at a single point in time. Am J Psychiatry. 1998;155:90-97.
If uncorrelated in twin pairs, unreliability of measure- 7. Torgersen S. The nature and origin of common phobic fears. Br J Psychiatry.
ment is indistinguishable from the effects of true indi- 1979;134:343-351.
8. Phillips K, Fulker DW, Rose RJ. Path analysis of seven fear factors in adult twin
vidual specific environment in our twin models. Our and sibling pairs and their parents. Genet Epidemiol. 1987;4:345-355.
short-term retest reliability found most tetrachoric cor- 9. Rose RJ, Ditto WB. A developmental-genetic analysis of common fears from early
relations for fears and phobias ranging from 0.60 to 0.85. adolescence to early adulthood. Child Dev. 1983;54:361-368.
These results suggest that a substantial proportion of what 10. Rose RJ, Miller JZ, Pogue-Geile MF, Cardwell GF. Twin-family studies of com-
we call E in our models in fact represents unreliability mon fears and phobias. In: Gedda L, Parisi P, Nance WE, eds. Intelligence, Per-
sonality, and Development. New York, NY: Alan R Liss Inc; 1981:169-174. Twin
of measurement. If our twin models were corrected for Research; vol 3, part 69B.
this effect, as shown in the female sample,17 the esti- 11. Neale MC, Fulker DW. A bivariate path analysis of fear data on twins and their
mates of the heritability of liability to phobias (ie, the ge- parents. Acta Genet Med Gemellol (Roma). 1984;33:273-286.